Posters
Friday/Saturday, 20-21 September 2002 $243
before and after RT. Materials and Methods: The target volumes of the elective and boost vol- Methods: Taste acuity tests and taste questionnaires were performed in 73 umes and the organs at risk -parotid glands, spinal cord, brain- were delin- patients with a head and neck cancer, treated with RT. The patients were eared on planning CT scabs of 12 patients. The cranial border of the lymph divided into 4 groups. Group 1 (n=17) was analyzed before the start of RT. node regions was delineated up to the cervical vertrebrael (C1), and up to Group 2 (n=17), group 3 (n=l 7) and group 4 (n=22) were at 2, 6 and t2-24 the cervical vertrebrae2 (C2), which is in general the surgical border in case months after RT, respectively. A cross-sectional analysis was performed of neck dissection, resulting in two elective volumes. The CRT and IMRT
between the 4 groups, planning was performed using a commercial system (PLATO RTS 2.5 and
Results: In group 1 (tested prior to RT), partial taste loss was observed in ITP 1.0). Prescribed dose to the elective volumes was 50Gy in 25 fractions 35, 18, 6 and 0% of patients for bitter, salt, sweet and sour, respectively. In and to the boost volume 70Gy in 35 fractions for CRT plans and the bio- group 2 (2 months after RT), taste loss (partial to total) was seen in 88, 82, logically equivalent dose in 30 fractions for IMRT plans. Normal tissue com- 76 and 53% for bitter, salt, sweet and sour, respectively. In group 3 (6 plication probability (NTCP) for xerostomia one year after radiotherapy was months after RT), partial taste loss was seen in 71, 65, 41 and 41% (bitter, calculated, using the parotid mean dose.
salt, sweet, sour) and in group 4 (12-24 months after RT) in 41, 50, 27 and Results: When irradiating level II nodes up to C2, the mean dose to the 27% (bitter, salt, sweet, sour). The difference in the incidence of taste loss parotid glands is decreased for both techniques, compared to irradiation up between the 4 groups was statistically significant (p<0.001 ; KruskaI-Wallis to C1. Results for the contralateral parotid gland are:
test). Distress caused by taste loss was most pronounced in group 2 (82%). Mean Dose (Gy)(sd) NTCP(%)(sd) Conclusions: Different tastes are lost after radiotherapy during different time C1 C2 C1 C2 periods. Bitter and salt qualities are most impaired. The sweet taste shows CRT 52(7) 49(7) 76(13) 69(18) the quickest recovery after RT. Gradual recovery is seen during the first tMRT 34(6) 27(5) 41(12) 25(7) year after treatment. Partial taste loss still persists 1-2 years after RT and is For the ipsilateral parotid glands relatively similar results are obtained but responsible for slight to moderate discomfort, the mean dose is higher due to the vicinity of the boost volume. For irradi- ation up to C1, it is 62Gy for CRT and it is reduced to 38Gy for IMRT. NTCP
817 Poster for xerostomia decreased approximately 7% and 16% in absolute values for
Trismus as a presenting symptom in nasopharyngeal carci-
CRT and IMRT plans respectively. Comparing the two techniques, the dosi- n o m a :its prognostic and response to treatment
metric advantage of IMRT is illustrated by a relative NTCP reduction up toE..(~zyar, M. Cengiz, M. Gurkaynak, I.L Atahan
64% for the same irradiated lymph node region.Department of Radiation Oncology, Hacettepe University, Ankara, Turkey
Conclusions: Lowering the cranial border of the level II lymph node regions to C2 could be consider, especially in IMRT techniques, since NTCP for Background: nasopharyngeal carcinoma (NPC) patients presenting with ini- xerostomia is reduced up to 70% relatively to conformal irradiation up to C1. tial trismus, were analysed in terms of general characteristics, response oftrismus to the treatment and prognostic role of it. 819 Poster
Methods: between 1993-2001, 232 patients (166 male, 66 female) with
Chemoradiation alone controls the
m a j o r i t y o fneck nodes
NPC were treated. Their ages ranged from 9 to 82 years. Histopathologi- i n c l u d i n g b u l k ydisease
cal diagnosis was WHO type II & III in 214 (91.4%) patients. All patients
A. Villar, J. Martinez, C. Fuentes, M. Espiheira, M. Perez, R. Hemandez,
were treated with definitive radiotherapy (XRT) and + chemotherapy.J. De Serdio, J. Artazkoz, J. Gil, J. Saavedra
Patient charts were retrospectively analysed and all initial symptoms, signs
Hospital N.S. Candelaria, Radiation Oncology, Santa Cruz de Tenerife,
and physical findings were documented. Of these 232 patients, it wasSpain
found that"20 (8.6%) patients had trismus, dental gap narrower than 3 cm,
at initial diagnosis. These patients were analysed for response of trismus Purpose:To evaluate the rate of neck nodes control achieved by primary to the treatment and prognostic role of trismus. Median follow-up time for chemoradiation
patients without trismus was 29 months (Range: 12-95 months) and for Patients and methods: From March 1996 to April 2002 64 patients with patients with trismus was 58 months (Range: 5-85 months) for surviving head and neck cancer (9 stage 3, 55 stage 4) were treated by a chemora-
patients, diation regimen already described ( Int J Rad Oncol Biol Phys 2001,50 (1):
Results: there were significantly more patients with young age (p<0.0001) 47-53). The regimen was composed of 2 fractions a day. Each fraction con- and advanced T status (p<0.0001) in patients with trismus. Overall 3 sisted of 115 cGy+ 5mg/m = of carboplatin with carbogen breathing. Treat- (37.5%) out of 8 patients were responded to the neoadjuvant chemothera- ment was given 5 days a week up to total doses of 350 mg/m 2 of carbo- py. Trismus was partially responded in 1 (12.5%) and completely respond- platin+8050 cGy in 7 weeks. Five patients presented with N0,t0 with N1,7 ed in 2 (25%) patients, overall response rate after chemotherapy and XRT with N2A,14 with N2B, 12 with N2C and 16 with N3.Bulky disease was very was found to be 88.2%. We found that 10 (90.9%) out of 11 patients with common (node dimension up to 16 cm of diameter). In one patient a neck complete response at the completion of treatment were alive with no evi- dissection was performed after completing chemoradiation for a node rest. dence of disease at their last control. It was found that 9 (80%) out of 11 The rest of patients were treated with chemoradiation only.
patients with complete response after treatment were trismus free at their Results:For a median follow-up of 44 months, actuarial Iocoregional control, last follow-up. 3 year Ioco-regional relapse free survival rates were 85.1% cause-specific survival and overall survival were 84%, 67% and 58% and 67.5%, for patients with and without trismus, respectively (p=0.66). our respectively. Actuarial node control for chemoradiation alone was 97%. univariate analysis have shown male gender and advanced nodal status to Including the case of neck surgery, actuarial node control was 98%. No be the significant unfavorable factors for OS, LRRFS & DMFS and after node recurrence has been seen more than 6 months after therapy. multivariat e analysis advanced nodal status remained as the single unfa- Conclusion: Chemoradiation alone is able to control de majority of neck vorable independent prognostic factor for all endpoints analysed. No sig- nodes, including bulky disease, provided that an appropriate regimen is nificant prognostic role of trismus was observed in both uni- and multivari- used. This is in contrast with a common opinion according to which, clini-
ate analysis, calty positive neck nodes, especially bulky disease, must be treated with
Conclusions: trismus is a more common symptom in NPC patients with surgery even if prior chemoradiation has been employed. The regimen young age and an indicator of advanced primary tumour. Trismus recov- described here is among the most active treatment schedules for head and ered in majority of patients at the end of treatment and patients with com- neck cancer, and, possibly has achieved the highest rate of node control plete recovery of trismus may have a better survival. However, no prog- including all surgical series
nostic significant role of trismus on survival was found in uni- and
multivariate analysis. 820 Poster
Decreased local control following radiation therapy alone in
818
Poster
early larynx cancer with anterior commisure extension
The influence of the cranial border of the level II lymph node
M. Ozsahin 1, L. Bron 2, P. Coucke 1, R.O. Mirimanoff 1, A. Zouhair 1
regions on xerostomia, for conformal and intensity-modular-
1Lausanne University Medical Center (CHUV), Radiation Oncology, Lau-
edirradiation of oropharyngeal cancer
sanne, Switzerland
E. Astreinidou, H. Dehnad, C.H.J. Terhaard, C.P.J. Raaijmakers
University Medical Center Utrecht, Radiotherapy, Utrecht, The Netherlands
2Lausanne University Medical Center (CHUV), Otorhinolaryngology, Lau-
sanne, Switzerland
Purpose: To investigate the influence of the cranial border of the level II
lymph node regions when irradiated electively on xerostomia, in case of Objective: To assess the patterns of failure in the treatment of early stage oropharyngeal cancer, for both conformal and intensity-modulated radiation squamous-cell carcinoma of the glottic larynx.
$244 Friday/Saturday, 20-21 September 2002
Posters
treated for laryngeal cancer (UICC T1N0 and T2N0) by radiation therapy 822 Poster
(RT) alone were retrospectively studied. Male to female ratio was 106/16,
Radiotherapeutic management of patients with head and
and median age 62 years (35-92). There were 68 patients with Tla, 18 withneck cancer, the influence
o f j o i nup change on dose unifor-
Tlb,
and 36 with T2 tumors. Diagnosis was done by biopsy in 104 patients,mity across photon field junctions by means of matching
h a l f laser vaporization or stripping in 18 Treatment planning consisted of 3D b l o c k f i e l d susing independent
j a w sconformal RT in 49 (40%) patients including 9 patientsirradiated using ary-
tenoid protection. A median dose of 70 Gy (60-74) was given (2 Gy/fr) over
C.B. Kwok 2, S- El-Saved 1, G. Lam 2
a median period of 46 days (21-79). Median follow-up period was 85
lUniversity of Ottawa, Radiation.Oncology, Ottawa, Canada
months.
2University of Ottawa, Medical physics, Ottawa, Canada
Results: The 5-year overall, cancer-specific, and disease-free survival was
respectively 80%, 94%, and 70%. Five-year local control was 83%. Median Independent collimation conveniently allows for the junctioning of abutting time to local recurrence in 19 patients was 13 months (5-58). Salvage treat- fields with non-diverging beam edges leading to a theoretical dosimetric ment consisted of surgery in 17 patients (one patient refused salvage and match. The actual dosimetry, at the central match plane as well as the influ- one was inoperable; total laryngectomy in t t, and partial laryngectomy or once of join up changes, has been tested in this study.
cordectomy in 6 patients). Six patients died because of laryngeal cancer Dose uniformity modeling was performed using the Therapran Plus version (ultimate larynx preservation rate: 95%). Univariate analyses revealed that 3.5 treatment planning system (TPS). Using a unit density phantom, two prognostic factors negatively influencing the local control were anterior types of modeling were performed: In the first, two half fields were abutted commissure extension, arytenoid protection, and total RT dose (< 66 Gy). at one of the two principal axes, One of the fields was kept constant and the Among the factors analyzed, multivariate analysis (Cox model) showed that other varied at 1 mm steps. In the second,the junction were moved once or anterior commissure extension, arytenoid protection, and male gender were twice by 1 cm. Off-axis ratios (OARs) at various depths were obtained. the worse independent prognostic factors in terms of local control. Siemens Mevatron accelerator was used to verify the accuracy of th~ TPS Conclusion: For early stage laryngeal cancer, RT alone offers an excellent simulations using identical photon energies and setups. Films were outcome. In case of anterior commissure extension, surgery or higher RT scanned and in-house software was used to convert optical density to dose. doses are warranted. Because of high relapse risk, arytenoid protection In the first type of TPS simulation we have found that dose non-uniformity should net be attempted, to be a linear function of gap - overlap sizes with a slope of - 12 ± 2 % for
all depths evaluated. In the corresponding accelerator measurement the
821 Poster same slope was obtained. The off-set of the line is at 0% (in the TPS case)
Patients with tonsillar carcinoma: a comparison of stereotac-
and ranging from a minimum of - 4 + 1% to a maximum of - 14 + 1 % fortic intensity modulated radiation therapy with brachytherapy
five Mevatron beams evaluated. In the second type of TPS simulation weK. Muller 1, P.J.C.M. Nowak 2, C. De Pan 2, A. Van Nimwegen 2, L. Lutharf 2,
have found that the dose uniformity improved by 2-fold when the junction isJ.C. Stroom 2, P.C. Levendag 2
moved once and by 3-fold when moved twice.The offsetting of the linear function in the case of measurement revealed a
1Erasmus Medical Center Rotterdam, Radiation Oncology, Rotterdam, The
systemic jaw overlaps with all the accelerators, well within the jaw calibra-Netherlands
lion tolerance limits. This error could not be improved further as 1 mm was2Erasmus Medical Center Rotterdam, Radiotherapy, Rotterdam, The Ne-
the smallest possible digital readout value. On the other hand an under-dos-therlands
ing of -14 + 1% was unacceptable.Dosimetric evaluation of abutted junction reveal that dose non-uniformity Introduction: Meta-analyses (Pignon, 2000) for cancers in the H&N have can be as high as - 14%. It is possible that the error can be corrected with shown that radiation therapy (RT) and concomitant chemotherapy result in collimator jaw calibration but this combine jaw error is within machine spec- improved tumor control at the cost of more toxicity. Thus it becomes of para- ification and quality assurance standard. The non-uniformity can be mount importance to reduce toxicity by implementing optimal sparing RT reduced using a small "gap". The required "gap" size can easily be deter- techniques. Cancer of the tonsillar fossa was taken as a model. In Rotter- mined using film measurement Out of the six linacs evaluated, five of them dam, external beam radiotherapy (ERT) followed by brachytherapy (BT) is require 1 to 2 mm gaps and 1 requires 0 to 1 mm gap for optimal field routinely implemented in T1-2(3) tumors with excellent clinical results. One matching. Further improvement in dose uniformity can be achieved by mov- way of optimizing the treatment is by sparing normal tissues, implementing ing the junction ± 1 cm within a fraction or between fractions.
IMRT from the start.
This paper focuses on the boost: BT is compared with 3DCRT, IMRT, 823 Poster
stereotactic radiation therapy (SRT), and IMRT combined with SRT (S-
IMRT for head and neck cancer: comparison of different opti-
IMRT).
mized irradiation techniques and first results of a QA pro-
Methods
and Material: For 9 treated NO patients with tonsillar fossa carol- g r a mnoma, a pro-treatment CT and a CT scan after implantation of atterloading
M. TomseL A. Margoum, V. Gr~goire, S. Vynckier, P, Scalliet
catheters was obtained. The clinical target volume (CTV) and critical struc-
U.C.L.-St-Luc University Hospital, Radiation oncology department, Brus-
lures (NT) were contoured on the pro-treatment CT. For the boost, asels, Belgium
3DCRT, an IMRT (Cadplan - Helios module) with MLC, a SRT and a S-
IMRT treatment plan (XPlan 2.0 and X knife RT 2.0 I~ 2) with mini MLC were The main goal of modulated-beam conformal therapy is to reduce the dose generated. The CTV to planning target volume (PTV) margin was 5 mm for to healthy tissue and sensitive structures around a uniformly irradiated tar- 3DCRT and IMRT, and 2 mm for SRT and S-IMRT. Using PLATO BPS ver- get volume. This concept is even more complicated in head and neck can- sion 14.2, the original BT treatment plan was reconstructed on CT. NT con- cers where several challenges occur for the optimization algorithm (i.e.
the
tours were copied to the post-implantation CT scans, and subsequently the complicated curvature of the patient external contour, the inhomogeneities dose in critical structures was computed, structures inside the patient, the concave shape of the planning target vol- Results: The median dose to the ipsilateral parotid gland was 9.8% (BT) ume around the spinal cord and the adjacent critical organs such as the and ranged from 14.1% (S-IMRT) to 58.3% (3DCRT) for ERT. For minor spinal cord and parotids). As a result, complex treatment plans have to be salivary glands (mucosa) the median dose was 17.1% for BT and varied achieved with a total dose of 70 Gy to be delivered to the planning target from 19.6% (IMRT) to 29.5% (3DCRT) for ERT. The dose contribution to volume and strict constraints had to be respected concerning the organs at other critical structures will be discussed at the meeting, risk.Conclusion: Depending on institutional policy, equipment and expertise, Different treatment plans were obtained with the MDS Nordion-TMS version one can opt for BT or S-IMRT when boosting T1-2(3) NO tonsillar carcino- 6.0.2 treatment planning system using intensity modulated fields of 6 MV mas. Besides specific advantages and limitations, with both modalities opti- from an ELEKTA linear accelerator, and intensity maps will be delivered mal sparing can be achieved. Some of the disadvantages with regard to tar- employing the step and shoot technique.
get, can be partly eliminated by using either CT/MRI treatment planning The aim of this study is then to analyze the different IMRT plans proposed (e.g. with BT) or placing tumor markers (e.g. with S-IMRT), Finally, for spar- for the head and neck cancers as a function of the different ballistics pro- ing major and minor salivary glands, it remains pivotal to implement IMRT posed, i.e. different combinations of intensity modulated fields, different in the first series of ERT (neck and primary), organs at risk with different constraints to their DVH in terms of isodose dis-
tributions and dose volume histograms.
This is conducting to a typical intensity modulated radiotherapy treatment of advanced head and neck cancers that will be specific for our department. An optimized intensity modulated treatment solution is then proposed and will be validated through a complete quality assurance program. Results of
